Geometrical stacking for fluid contact analysis Examples from the ...

Geometrical stacking for fluid
contact analysis
Examples from the North Sea
Dutch sector
* Suncor Netherlands
Arnaud Huck
Frits Blom*
January 14, 2010

Outline
Introduction
Base Vlieland Shale example
Top Middle Bunter example
Top Rotliegend example
Conclusions

Outline
Introduction
Base Vlieland Shale example
Top Middle Bunter example
Top Rotliegend example
Conclusions

Common Contour Binning
Objective:
– Detect subtle hydro-carbon related anomalies and pin-point contacts
(GWC, GOC and OWC)
Principle:
– Seismic traces that penetrate the top reservoir at the same depth (i.e.
lie on the same contour line) have identical hydro-carbon columns.
– Stacking traces along the same contour line thus enhances possible
hydrocarbon effects while stratigraphic variations and noise are
cancelled.
CCB Outputs:
– New 3D volume with stacked traces along contour lines
– CCB stack = 2D section with stacked traces
– Crossplot of stacked amplitudes vs. bin depth
Developed by and for

CCB plugin principle
polygon outlines
area of investigation
Amplitudes
Seismic
at top
reservoir Stacked Amplitudes
traces
after traces at
top will
CCB
be
inside
Top reservoir
polygon are
map
stacking
stacked along
re-distributed along
contour lines
contour lines
Contour line
Top reservoir

Outline
Introduction
Base Vlieland Shale example
Top Middle Bunter example
Top Rotliegend example
Conclusions

Example 1: Base Vlieland Shale
In this first example Common contour binning was applied
around the Base Cretaceous Unconformity (Base Vlieland
Shale) at a known prospect.
The formation below this unconformity is a good reservoir
containing gas, while the formation above is a regional
seal.

Base Vlieland Shale depth map
Contours every 5ms
East-west size:
2.6 km
TWT in s.

Seismic amplitude map
at Base Vlieland Shale
A large low
amplitude area is
visible in the
northern part
It does not
correspond
completely with the
time structure map
A second larger
area show a quick
amplitude decrease
with polarity
reversal
Seismic amp.

CCB Stack, Area = entire survey
Base Vlieland Shale
The following observations
can be made:
1465ms: Drop of amplitude
by 20%
Low amplitudes between
1465 and 1475ms
Much larger drop of
amplitudes between 1482 and
1492ms
Besides the small changes
around 1465 ms an enormous
amplitude change with phase
reversal is observed.
This change is not step-like
but the inflexion point is
located at 1486 ms. The
change occurs within +/-2ms.

The large
amplitude change
is marked as the
limit between the
blue and green
areas
CCB Stack amplitude map
Base Vlieland

Conclusions
Base Vlieland Shale
The CCB stack pattern features two amplitude anomalies at
1465 and 1486ms.
The amplitude change at 1465 ms is very subtle and
cannot be confirmed on time data.
The large amplitude anomaly at 1486 ms is considered
typical behaviour for a CCB stack and could be interpreted
as the position of the expected gas/water contact.
This fluid contact has been confirmed by drilling: Prior to
drilling the flat spot was estimated at 1496ms. The GWC in
the well was actually found at 1489ms, much closer to the
CCB estimation.

Outline
Introduction
Base Vlieland Shale example
Top Middle Bunter example
Top Rotliegend example
Conclusions

Top Middle Bunter
Contours every 25ms
TWT in s.

Seismic amplitude map
at Top Middle Bunter
The top of the
structure features
high amplitudes
Seismic amp.

2D CCB Stack
Top Middle Bunter
The analysis was run in and around
the prospect. Its limit is around a
bin depth of 1508ms.
The CCB stack reveals a trend of
decreasing amplitudes wrt bin
depth.
Nevertheless no clear amplitude
change is visible.

Conclusions
Top Middle Bunter
The CCB stack at Top Middle Bunter shows smooth
variations in amplitudes, somewhat decreasing with bin
depth.
This behaviour does not correspond to a saturation change
in a formation whose thickness is above tuning resolution.
Decreasing/increasing amplitude with bin depth can be
related to both thickness and saturation changes. Only
forward modelling can distinguish them, hence common
contour binning is not conclusive at Top Middle Bunter.

Outline
Introduction
Base Vlieland Shale example
Top Middle Bunter example
Top Rotliegend example
Conclusions

Seismic data,
random line through prospect
5 km 2 (Rotliegend prospect)
The Middle Bunter which is
between Vlieland and
Rotliegend is completely eroded
at the top of the structure.
Base Vlieland
Top Rotliegend

Top Rotliegend
The red polygon is
delimiting the area of
interest
Contours every 25ms
East-west size
(prospect): 10 km
Two areas are visible
within the prospect,
clearly separated by
a normal fault with a
70ms offset at top
Rotliegend.
TWT in s.
TWT in
sec.

Seismic amplitude map
at Top Rotliegend
Several blocks are
visible within the
prospect.
The crest features
lower amplitudes
than the southeastern
part.
Seismic amp.
TWT in
sec.

Isopach
Top Rotliegend - Base Vlieland Shale
Several blocks are
visible within the
prospect.
Zero thickness
at crest
The high amplitudes
are located in the
part where there is
an intermediate
formation between
Vlieland and
Rotliegend, around
40ms TWT in
thickness.
Middle Bunter
thickness:
40+/-10ms
Time difference in ms.
TWT in
sec.

CCB Stack, Area = all
Top Rotliegend
The bin depths outside
(thus deeper than) the
prospect show again low
amplitudes.
The outer part of the
prospect and observed
amplitude change is at
1464ms TWT.

CCB Stack amplitude map
Top Rotliegend
CCB Stack combined from the 3 sub stacks
Deep areas:
Low amplitudes
Top of the structure:
Low amplitudes
Second structure:
High amplitudes

Seismic amplitude map
at Top Rotliegend
Several blocks are
visible within the
prospect.
The crest features
lower amplitudes
than the southeastern
part.
Seismic amp.
TWT in
sec.

Pre-stack CCB
Top Rotliegend
CCB was applied to prestack
data. The technique
works similarly with offset
as extra dimension.
The display is limited to
the horizon Z only.
The vertical axis
represent offsets (in m.)
while the colours remain
seismic amplitudes.
A classical AVO graph
would be a vertical slice
on that panel.

Conclusions
Top Rotliegend
CCB stacking highlighted clear amplitude changes at 1406 and
1464ms TWT, from the post and pre-stack data.
The structure is divided in two by a normal fault that appears to
be a lateral seal, since high amplitudes are observed in the south
east compartment that are not present in the crest, neither at the
same depth nor shallower.
Therefore three areas may be distinguished:
– The crest is a Rotliegend/Vlieland contact, with low amplitudes
– The outer part of the prospect is a Rotliegend/Bunter contact
with low amplitudes.
– The south-eastern part of the prospect is a Rotliegend/Bunter
contact with much higher amplitudes.

Conclusions
Top Rotliegend
While the amplitude change at 1406ms could be attributed
to a lithology change rather than to a fluid contact, the
second amplitude change at 1464ms cannot be explained
by a lithology change based on the available interpretations
and the observed structures.
Therefore the CCB results and structural information
indicate that the south-eastern part of the prospect is gas
filled with an up-dip seal along the normal fault in the
north-west, and a gas-water contact at 1464ms +/-4ms.

Outline
Introduction
Base Vlieland Shale example
Top Middle Bunter example
Top Rotliegend example
Conclusions

Conclusions
Common contour binning highlights subtle fluid-related seismic anomalies
and allows delineating fluid contacts with more accuracy. However, the
technique in isolation cannot uniquely prove the presence (or absence) of
hydrocarbons.
The first and last example showed how gas water contacts can be picked
with confidence, at Base Vlieland Shale and Top Rotliegend, respectively.
However, the second example shows also how difficult the interpretation
can be when the amplitude increases or decreases monotonously with bin
depth. Such difficult cases cannot be interpreted with merely the
application of common contour binning. They require further study
including additional structural information and sensitivity analyses.
Furthermore, there is a significant risk at using this geometrical stacking
technique outside of the original depth domain, therefore it is always
recommended to go in the depth domain when possible.